Counteracting effects of urea and betaine in mammalian cells in culture PAUL H. YANCEY AND MAURICE B. BURG Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung, and Blood Institute, Bethesda, Maryland 20892 YANCEY, PAUL H., AND MAURICE B. BURG. Counteracting effects of urea and betaine in mammalian cells in culture. Am. J. Physiol. 258 (Regulatory Integrative Comp. Physiol. 27): R198-R204, 1990.-Urea and methylamines, such as betaine, are among the major organic osmotic effecters accumulated by organisms under hyperosmotic (high NaCl) stress; the mam- malian renal medulla also accumulates such compounds in antidiuresis. Studies on isolated proteins show that urea gen- erally destabilizes protein structure, whereas methylamines are generally stabilizers capable of offsetting the effects of urea. The counteracting-osmolytes hypothesis predicts that cells ex- posed to high urea concentrations require methylamines for optimal function. In this study, urea, betaine, and other solutes (NaCl, glycerol, sorbitol) were added to growth medium of cultured mammalian cells under conditions in which most solutes entered the cells. For two renal [Madin-Darby canine kidney (MDCK) and PAP-HT25] and one nonrenal (Chinese hamster ovary) cell line, urea (X00 mM) or betaine (X0-100 mM) alone inhibited cell growth and survival, measured as colony-forming efficiency. However, the addition of betaine (up to 120 mM) to media with urea (50-300 mM) greatly increased colony-forming efficiency above that with urea alone. A similar, although less marked effect, was seen on colony sizes with MDCK cells. These results support the counteracting-osmo- lytes hypothesis. Madin-Darby canine kidney cells; PAP-HT25; Chinese ham- ster ovary cells; sodium chloride; osmoregulation; glycerol; sor- bitol; cloning efficiency; organic osmolytes WHEN EXPOSED to hyperosmotic stress, cells of a wide variety of organisms, from eubacteria to mammals, main- tain osmotic balance by accumulating certain organic solutes rather than inorganic ions. These organic osmo- lytes comprise three classes of compounds (23), namely 1) sugar alcohols or polyols such as glycerol (e.g., in halotolerant algae, freeze-tolerant insects), 2) neutral free amino acids and related solutes (e.g., in marine invertebrates), and 3) urea in combination with methyl- amines (e.g., in marine cartilaginous fishes). Two hypotheses attempt to explain these patterns. The first is the “compatible-osmolytes” hypothesis. In numerous studies on isolated protein systems, polyols and certain amino acids, unlike NaCl or KCl, do not significantly perturb protein function over a wide range of concentrations (6, 8, 24). The compatible osmolytes hypothesis (6, 8) predicts that such solutes can be safely accumulated to high concentrations by cells without significant effects on cell activities and should be com- R198 monly found in osmotically stressed organisms. The second or “counteracting osmolytes” hypothesis is based on very different effects of urea and methyla- mine compounds such as trimethylamine oxide and be- taine (N-trimethylglycine). Urea is well known to be a potent destabilizer of protein structure and generally an inhibitor of function, i.e., it is not a compatible solute. However, methylamines have been found to be effective stabilizers of protein structure (e.g., thermal stability) and generally activators of functional properties [Km, maximal velocity ( Vmax)]. When combined with urea at about a 1:2 methylamine-to-urea concentration ratio, their effects can often offset those of urea. This counter- acting mechanism was first revealed in studies with proteins of elasmobranch fishes, whose cells contain methylamines and urea at about this l:2. This mecha- nism has also been demonstrated for a variety of isolated, nonliving systems, including measurements of enzyme kinetics, stability of elasmobranch and mammalian pro- teins, and contraction of demembranated shark muscle fibers (23, 24). Thus the counteracting osmolytes hy- pothesis predicts that perturbations of cell functions will be minimal at a 1:2 methylamine-urea concentration, and such combinations should generally be found in urea- based osmoregulatory systems (24). Both the compatible and the counteracting hypotheses are based largely on observations of osmolyte patterns in organisms and experiments with nonliving systems (23). Both make specific predictions about the effects of such compounds in living cells, but tests of these are rare, especially with those of higher eukaryotes. A few studies have tested the effects of these osmolytes in living microorganisms; for example, a methylamine (betaine) has been shown to enhance survival of bacteria in hy- perosmotic (NaCl) solutions (7, 9). But no study with living cells or tissues has directly tested the proposed relation between methylamines and urea. It is important to do so, particularly because the simple counteracting effects of urea and methylamines seen with proteins may be more complex in whole cells. First, for a few urea- inhibited proteins, no opposing effects of methylamines were seen (24, 27). Thus counteraction might not be effective over the integrated activities of a whole cell. Second, although urea generally inhibits and methyla- mines generally activate isolated protein systems, for one enzyme studied, counteraction occurred with urea as an activator and a methylamine as an inhibitor (27). Also, the increased Vmax or decreased K, values of enzymes by 10.220.33.2 on April 5, 2017 http://ajpregu.physiology.org/ Downloaded from
Transcript
Counteracting effects of urea and betaine in mammalian cells in culture
PAUL H. YANCEY AND MAURICE B. BURG Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung, and Blood Institute, Bethesda, Maryland 20892
YANCEY, PAUL H., AND MAURICE B. BURG. Counteracting effects of urea and betaine in mammalian cells in culture. Am. J. Physiol. 258 (Regulatory Integrative Comp. Physiol. 27): R198-R204, 1990.-Urea and methylamines, such as betaine, are among the major organic osmotic effecters accumulated by organisms under hyperosmotic (high NaCl) stress; the mam- malian renal medulla also accumulates such compounds in antidiuresis. Studies on isolated proteins show that urea gen- erally destabilizes protein structure, whereas methylamines are generally stabilizers capable of offsetting the effects of urea. The counteracting-osmolytes hypothesis predicts that cells ex- posed to high urea concentrations require methylamines for optimal function. In this study, urea, betaine, and other solutes (NaCl, glycerol, sorbitol) were added to growth medium of cultured mammalian cells under conditions in which most solutes entered the cells. For two renal [Madin-Darby canine kidney (MDCK) and PAP-HT25] and one nonrenal (Chinese hamster ovary) cell line, urea (X00 mM) or betaine (X0-100 mM) alone inhibited cell growth and survival, measured as colony-forming efficiency. However, the addition of betaine (up to 120 mM) to media with urea (50-300 mM) greatly increased colony-forming efficiency above that with urea alone. A similar, although less marked effect, was seen on colony sizes with MDCK cells. These results support the counteracting-osmo- lytes hypothesis.
WHEN EXPOSED to hyperosmotic stress, cells of a wide variety of organisms, from eubacteria to mammals, main- tain osmotic balance by accumulating certain organic solutes rather than inorganic ions. These organic osmo- lytes comprise three classes of compounds (23), namely 1) sugar alcohols or polyols such as glycerol (e.g., in halotolerant algae, freeze-tolerant insects), 2) neutral free amino acids and related solutes (e.g., in marine invertebrates), and 3) urea in combination with methyl- amines (e.g., in marine cartilaginous fishes).
Two hypotheses attempt to explain these patterns. The first is the “compatible-osmolytes” hypothesis. In numerous studies on isolated protein systems, polyols and certain amino acids, unlike NaCl or KCl, do not significantly perturb protein function over a wide range of concentrations (6, 8, 24). The compatible osmolytes hypothesis (6, 8) predicts that such solutes can be safely accumulated to high concentrations by cells without significant effects on cell activities and should be com- R198
monly found in osmotically stressed organisms. The second or “counteracting osmolytes” hypothesis
is based on very different effects of urea and methyla- mine compounds such as trimethylamine oxide and be- taine (N-trimethylglycine). Urea is well known to be a potent destabilizer of protein structure and generally an inhibitor of function, i.e., it is not a compatible solute. However, methylamines have been found to be effective stabilizers of protein structure (e.g., thermal stability) and generally activators of functional properties [Km, maximal velocity ( Vmax)]. When combined with urea at about a 1:2 methylamine-to-urea concentration ratio, their effects can often offset those of urea. This counter- acting mechanism was first revealed in studies with proteins of elasmobranch fishes, whose cells contain methylamines and urea at about this l:2. This mecha- nism has also been demonstrated for a variety of isolated, nonliving systems, including measurements of enzyme kinetics, stability of elasmobranch and mammalian pro- teins, and contraction of demembranated shark muscle fibers (23, 24). Thus the counteracting osmolytes hy- pothesis predicts that perturbations of cell functions will be minimal at a 1:2 methylamine-urea concentration, and such combinations should generally be found in urea- based osmoregulatory systems (24).
Both the compatible and the counteracting hypotheses are based largely on observations of osmolyte patterns in organisms and experiments with nonliving systems (23). Both make specific predictions about the effects of such compounds in living cells, but tests of these are rare, especially with those of higher eukaryotes. A few studies have tested the effects of these osmolytes in living microorganisms; for example, a methylamine (betaine) has been shown to enhance survival of bacteria in hy- perosmotic (NaCl) solutions (7, 9). But no study with living cells or tissues has directly tested the proposed relation between methylamines and urea. It is important to do so, particularly because the simple counteracting effects of urea and methylamines seen with proteins may be more complex in whole cells. First, for a few urea- inhibited proteins, no opposing effects of methylamines were seen (24, 27). Thus counteraction might not be effective over the integrated activities of a whole cell. Second, although urea generally inhibits and methyla- mines generally activate isolated protein systems, for one enzyme studied, counteraction occurred with urea as an activator and a methylamine as an inhibitor (27). Also, the increased Vmax or decreased K, values of enzymes
OPPOSING EFFECTS OF UREA AND BETAINE IN CELLS R199
usually found with methylamines could be detrimental to complex regulatory interactions within cells. Thus either urea or methylamines could be inhibitory in the absence of the other in living cells, but effective counter- action could still occur at a l:2 methylamine-urea con- centration (23, 24).
Recently, an osmoregulatory system involving urea and methylamines has been found in the mammalian renal medulla. In antidiuresis, this tissue accumulates high concentrations of urea and NaCl, and in apparent response to this, cells of the medulla accumulate polyols, mainly sorbitol and myoinositol, and methylamines, mainly betaine and glycerophosphorylcholine (GPC) (2, 5, 25). The simultaneous occurrence of high NaCl and polyols on the one hand, and urea and the methylamines on the other, fulfills one prediction of both the compat- ible and counteracting osmolytes hypotheses but does not provide direct evidence for compatible and counter- acting properties with cell functions.
The existence of cell lines from mammalian kidneys has provided an opportunity for more direct tests of the roles of these compounds in the osmoregulation of living eukaryotic cells. Recently, a renal cell line (GRB-PAPl) was established that synthesizes intracellular sorbitol in response to elevated extracellular NaCl (3). Specific in- hibitors for the enzyme (aldose reductase) that synthe- sizes sorbitol are available. This has allowed a demon- stration that sorbitol production is directly correlated with cell growth and survival in hyperosmotic media (26), providing strong support for the compatible osmo- lytes hypothesis.
In the present studies, we sought a similar direct test for the counteracting osmolytes hypothesis, using estab- lished cell lines from kidneys and another mammalian tissue to examine the effects of urea and the methylamine betaine on cell growth. Lacking specific inhibitors for cellular accumulation of betaine, we added urea and betaine to growth media and found that these compounds readily penetrate the cells. Polyols and NaCl were also tested.
Colony-forming efficiency (also termed “plating” or “cloning” efficiency) was used as a measure of cell sur- vival and growth. We examined the effects of high con- centrations of various solutes on the colony-forming efficiency of three lines of mammalian cells in culture. 1) Madin-Darby canine kidney (MDCK) cells are a line of epithelial cells from dog kidney. When grown in hy- perosmotic medium, they accumulate betaine, GPC, and myoinositol as osmolytes (15). 2) PAP-HT25 cells are a line of epithelial cells from rabbit renal inner medulla (21). They accumulate large amounts of sorbitol when grown in hyperosmotic medium (3). 3) Chinese hamster ovary cells (CHO) (10) were chosen to be neither renal nor epithelial. The results confirm the basic tenet of the counteracting osmolytes hypothesis, namely that high concentrations of urea alone decreased cell survival and growth, but the addition of betaine substantially reversed this deleterious effect of urea.
MATERIALS AND METHODS
Cell Culture
CHO cells were grown in a-modified minimal essential medium (NIH) supplemented with 10% fetal bovine
serum, 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin (10). MDCK cells (passages 59-99) were generally grown in Dulbecco’s minimum essential medium (NIH) supplemented as for CHO cells. PAP- HT25 cells (passages 53-76) were grown in Coon’s mod- ified Ham’s F12 (NIH), supplemented as described pre- viously (21). Urea, betaine, NaCl, glycerol, and/or sor- bitol were added to these media as solids to the concen- trations indicated, followed by filtration sterilization. Osmolalities of all media were checked with a vapor- pressure osmometer (Wescor, Logan, UT).
Cell Osmolyte Contents
Ten-centimeter plates were seeded with MDCK cells that were then grown for 4-5 days to 90-95% confluence. The plates were washed twice with 7 ml of isotonic phosphate-buffered saline (PBS), scraped in 1.0 ml of PBS, and then scraped into 0.5 ml cold 7% perchloric acid. The extract was centrifuged for 10 min with a microcentrifuge. The pellets were dissolved by shaking overnight in 1 N NaOH. The resulting solution was measured for protein by the BioRad (Richmond, CA) Protein Assay (using gamma-globulin in 1 N NaOH as a standard). The supernatants were then neutralized with KOH, passed through lipid-removing Sep-Pak C18 car- tridges (Waters, Milford, MA) and 0.45 PM filters, and then analyzed for organic solutes by high-performance liquid chromatography (HPLC) with a Waters SugarPak column, following the method of Wolff et al. (22). Briefly, 50-~1 samples were injected into a mobile phase of 50 mg/l calcium disodium EDTA running at 0.6 ml/min through the column at 84OC. Compounds were detected by refractive index, and peak analysis was performed using Baseline software (Waters).
Colony-Forming Efficiency
Colony-forming efficiency is defined as number of successful colonies divided by number of cells seeded. These were determined by seeding -500 (CHO), 1,000 (MDCK), or 3,000 (PAP-HT25) cells (cell numbers de- termined by hemocytometer) on lo-cm plates in 8 ml medium. PAP-HT25 cells were allowed to settle 24 h in control medium before being switched to test medium. CHO cells were seeded directly in the test media to avoid the dislodging of cells during medium changes (10). MDCK cells were tested by both methods; colony-form- ing efficiencies differed by <lo% between the two, and the results reported are from experiments using the direct-seeding technique. The attached MDCK and PAP-HT25 cells were fed every 3-5 days with 7-8 ml medium. Colonies were allowed to grow to l-2 mm di- ameter, which required lo-15 days in control medium. A longer wait was necessary in some of the media that slowed cell growth. CHO colonies were grown for 6 days without feeding to avoid dislodging mitotic cells (lo), at which time they had reached l-2 mm diameters. Colonies were fixed in 10% Formalin in PBS (15 min), rinsed with distilled water, and stained in 1% toluidine blue 0 (15 min, followed by water rinse) or they were fixed and stained for 15 min in 50% ethanol with 0.5% methylene
RZOO OPPOSING EFFECTS OF UREA AND BETAINE IN CELLS
blue followed by water rinse. The number of colonies was then counted under a dissecting microscope, with sizes determined using an ocular micrometer.
Cyanate
Urea is known to break down slowly into ammonium (iso)cyanate, requiring several days at 37°C and several weeks in the cold to reach equilibrium (14). Cyanate is known to carbamylate the N-terminal of proteins, often changing their functions (20) and could explain any effects of urea on cell growth. To prevent cyanate accu- mulation, media were stored at 4°C and were changed in incubated plates every 3-4 days (except for CHO cells). In addition, contents of cyanate in media were analyzed by infrared spectroscopy in 0.025 mm calcium fluoride cells. NaOCN (3.5 mM) dissolved in water or in MDCK growth medium was used as a standard, giving a signal at 2,164 cm-‘. No cyanate could be detected in media with 225 mM urea incubated for 7 days at 37°C. As another test, cells were grown in 2.0 mM ammonium cyanate (2.0 mM NH&l plus 2.0 mM sodium cyanate), the equilibrium value predicted for 225 mM urea (14). This resulted in a 23% (+3 SD) reduction in colony- forming efficiency of MDCK cells, compared with the 76% (214) reduction seen in 225 mM urea (see below). When MDCK cells were grown with 2.0 mM ammonium cyanate and 50 mM betaine, colony-forming efficiency was no higher than with ammonium cyanate alone, in contrast with increased colony-forming efficiency when betaine was added to media containing urea (see RE- SULTS). Thus we conclude that cyanate effects, if any, were small and are qualitatively different from the effects of urea.
Statistics
After arcsin transformation for percentage data (col- ony-forming efficiencies), all data were analyzed by analysis of variance (ANOVA).
RESULTS
Cellular Contents of Organic Solutes
To determine whether the solutes added to the growth media were entering the cells, MDCK cells were grown for 4-5 days to a sufficient density for analysis (90-95% confluent). Osmolyte contents of washed and scraped cells were determined by HPLC. As expected, urea and glycerol penetrated the cells. In media with 200 mM of these solutes, cell contents were -1,100 mmol/mg cell protein (Table l), a ratio of 5.5. If the assumption is made that these small solutes were equilibrated across the plasma membrane (15) to 200 mM in cell water, then this ratio gives an estimate of cell protein contents of 180 mg/l cell water. Betaine in the medium was also penetrated (Table 1). Using the 5.5 ratio, we estimate that concentrations (mM) were as high as or higher than in the medium. For example, cells in medium with 50 mM betaine are estimated to have 80 mM betaine.
Na contents of cells were not measured, but previous studies with MDCK cells grown under the same condi-
TABLE 1. HPLC analyses of organic osmolytes in MDCK cells grown for 4-5 days to 90-95% confluence
Medium Osmolality Urea Betaine
Control 305 (0) (0) 200 mM Urea 490 1,194*39 (0) 50 mM Betaine 355 (0) 455.6t14.2 200 mM Urea, 50 mM betaine 540 1,130+46 365.6k30.5 200 mM Urea, 100 mM betaine 590 1,032+99 582.5t73.5 50 mM NaCl 390 (0) 34.4k12.1
Glycerol
200 mM Glycerol 555 1,195, 1,209
Sorbitol
50 mM Sorbitol 353 292k12 - Values are means t SD for 3 replicate plates unless indicated, in
mmol/kg cell protein. (0), none detectable.
tions showed that there is little sustained absolute in- crease in intracellular NaCl in hyperosmotic media (15).
Effects of Individual Solutes on Colony-Forming Efficiency
Tests of colony-forming efficiency were conducted in separate experiments under the same conditions used for osmolyte content determinations except that the cells were at a lower density.
Controls. In control media (without added test solutes), MDCK, PAP-HT25, and CHO cells had colony-forming efficiencies of 30-40%, 6-7%, and 60-70%, respectively. These values are taken as 100% for comparisons to test conditions (although not for statistical analysis).
Urea. At higher concentrations, this solute inhibited colony-forming efficiency of all three cells types, with PAP-HT25 cells being the most sensitive (Figs. lA, 2, 3). Below 150 mM, urea had no effect or was stimulatory with MDCK (Fig. IA) and CHO (Fig. 3) cells.
Betaine (N- trimethylglycine). This solute showed a small but significant stimulation of colony-forming effi- ciency with MDCK cells at 50 mM (Fig. 1A) but had little or slightly inhibitory effects on the other cell types at 50-100 mM (Figs. 2 and 3). There was a steep decline in colony-forming efficiency for MDCK cells above 100 mM betaine (Fig. lA); these higher concentrations were not tested with the other cell types.
NaCl. When added alone at 29.5 mM (50 mosM), NaCl had no effect on PAP-HT25 or CHO cells (Figs. 2 and 3) but greatly reduced the number of MDCK colonies (Fig. 1A).
Polyols. Glycerol caused a decrease in colony-forming efficiency of MDCK cells that was linear with concen- tration (or osmolality) (Fig. 1A ). At low concentrations, it was more inhibitory than urea or betaine, but at high concentrations, it was less so. Sorbitol, in contrast, in- hibited colony-forming efficiency at 50 mM to about the same degree as 25 mM NaCl, much more than did glyc- erol (Table 2). Other cell types were not tested.
Effects of Solute Combinations on Colony-Forming Efficiency
Urea plus betaine. For most combinations tested, the addition of 50 mM betaine to various concentrations of
FIG. 2. Colony-forming efficiency with PAP-HT25 cells with effects of betaine, NaCl, urea alone, and in combinations indicated. Values are for 3-4 replicate plates *SD. Increased efficiencies in betaine-urea combinations (dotted lines) were significant (P < 0.01, ANOVA) for all cases. * 2:1 urea-to-betaine ratio.
OSMOLALITY, moam/kg
t
NaCl
--,+a Add betaine
to wea
0 L 300 350 400 450 500 550 600 650
OSMOLALITY, mown/kg
FIG. 3. Colony-forming efficiency with Chinese hamster ovary (CHO) cells with effects of betaine, NaCl, urea alone, and in combi- nations indicated. Values are for 3-6 replicated plates *SD. Increased efficiencies in betaine-urea combinations (dotted lines) were significant (P < 0.03, ANOVA) for all but the 400-500 mosmol/kgHzO points. * 2:1 urea-to-betaine ratio.
NaCl . \
TABLE 2. Relative colony-forming efficiency of MDCK
A-%A 1 cells in serum-based media with various
I I I solutes added as indicated 300 350 400 450 500 550 600 650
OSMOLALITY, mosm/kg
FIG. 1. Colony-forming efficiency with MDCK cells. Values are means of 3-6 replicate plates, tSD (error bars not shown if within size of plotting symbol). A: effects of urea, betaine, glycerol, and NaCl alone. Increased efficiencies in 50 mM urea or betaine (350 mosmol/kgH20) were significant (P < 0.05, ANOVA). Decreases were significant (P c 0.05) for urea, betaine, glycerol, and NaCl at and above the 480, 400, 400, and 340 mosmol/kgHzO solutions, respectively. B: effects of urea and betaine alone or in combination as indicated. Increased efficiencies in betaine-urea combinations (dotted lines) were significant (P < 0.01, ANOVA) for all but 350-400 and 440-490 mosmol/kgH,O cases. * 2:1 urea-to-betaine ratio. C: effects of glycerol and NaCl alone and in combination with betaine as indicated. Efficiencies in betaine combi- nations (dotted lines) were significantly increased (P c 0.01, ANOVA) only for 340-390 mosmol/kgH20 NaCl case and significantly decreased (P C 0.01, ANOVA) only for 595-645 mosmol/kgHzO glycerol case.
urea significantly increased the colony-forming effi- NaCZ plus betaine. When NaCl-containing media were ciency of all cell types (Figs. lB, 2, 3). The addition of supplemented with 50 mM betaine, colony-forming effi- urea and betaine (up to 120 mM for MDCK cells) at a ciency of MDCK cells increased in 25 mM NaCl (43 2:l concentration (asterisks, Figs. 1B and 3) generally mosM added) but not in 50 mM NaCl (85 mosM) above
Non-Urea Solute Added
Colony-Forming Efficiency
%With solute alone %With 225 mM urea
None 100.0~3.0 33.9k2.7 50 mM Betaine 121.7*4.3* 67.1t9.3” 25 mM NaCl 39.8t8.4* 36.9t4.7 50 mM Glycerol 90.7t8.0 31.7k3.0 50 mM Sorbitol 37.4&5.7* 27.526.6
Values are means t SD for 3-6 replicas. Cells were seeded at low densities and grown to l-2 mm colonies. Efficiency for control condi- tion was taken as 100%. * Significant (P < 0.05, ANOVA) increase or decrease from “None” (100 or 33.9%).
increased colony-forming efficiency (compared with urea alone) even more at urea concentrations X00 mM, but the effect was not linear with betaine concentration.
R202 OPPOSING EFFECTS OF UREA AND BETAINE IN CELLS
that for NaCl only (Fig. 1C). Other cell types were not tested.
Glycerol plus bet&e. Betaine (50 mM) added to glyc- erol at two different concentrations did not alter or reduced colony-forming efficiency of MDCK cells com- pared with glycerol alone (Fig. 1C). Other cell types were not tested.
Urea plus polyols. When added to media with 225 mM urea, 50 mM glycerol or sorbitol showed no counteracting properties (Table 2).
Effects of Added Solutes on Colony Sizes
For some colony-forming efficiency experiments, sizes of MDCK colonies were tallied and found to follow approximately normal distributions, for which means t SE were calculated. Glycerol had the least effect on mean colony size, followed by urea, betaine, and NaCl (Fig. 4A ). Addition of betaine with urea resulted in increased colony size in some cases but not in others (Fig. 4B).
DISCUSSION
The experiments here provide the first tests of the counteracting solutes hypothesis with living eukaryotic
t 6 d 0.5
0
550 OSMOLALITY, mosm/kg
600 650
0.0 1 I I I 1 I I
300 350 500 550 OSMOLALITY, mosn/kg
FIG. 4. Colony sizes in MDCK cells, with means &SE for 3-6 replicate plates. A: effects of glycerol, urea, betaine, and NaCl alone. Decreased sizes were significant for all solutes at all points above control. B: effects of betaine added to urea. Increased efficiencies in betaine-urea combinations (dotted lines) were significant (P c 0.03, ANOVA) for all but 580-630 mosmol/kgHzO points. * 2:l urea-to- betaine ratio.
cells. The results demonstrate directly that a methyla- mine, betaine, can play a protective role against urea in a living cell, as predicted by that hypothesis. The physi- cochemical basis for the differing effects of compatible, stabilizing, and perturbing solutes is incompletely under- stood but appears to relate to binding or exclusion of solutes from backbones and immediate hydration spheres of macromolecules (e.g., proteins) (13). Because all of the solutes tested in the present study clearly penetrated MDCK cells (Table l), effects seen on colony-forming efficiencies may be caused by such interactions among the osmolytes and intracellular macromolecules (al- though of course that cannot be directly shown). The results could also be due to extracellular interactions, but as noted in MATERIALS AND METHODS, little differ- ence in initial attachment of cells to plates was seen between seeding in control or test media. In addition, some of the effects could be simply due to osmotic pressure differences created by addition of solutes to media; this possibility will be examined below.
Effects of Individual Solutes
Urea. The adverse effects of urea on colony-forming efficiency were found with two renal (MDCK, PAP- HT25) and one nonrenal (CHO) mammalian cell lines (Figs. l-3) at all concentrations X00 mM (Figs. l-3). For one renal line (MDCK cells), colony sizes were reduced at all concentrations (Fig. 4A). Such results are expected based on the widespread ability of urea to inhibit isolated macromolecular and complex physiolog- ical functions (23, 24). The effects are not likely to be caused by osmotic pressure differences, since urea is known to equilibrate rapidly across kidney cell mem- branes in culture (15).
At concentrations of loo-150 mM, urea did not inhibit colony formation by MDCK (renal) and CHO (non- renal) cells (Figs. 1A and 3) and in fact significantly increased survival of the former at 50 mM (Fig. 1A). In contrast, significant inhibition of PAP-HT25 (renal) cells was seen at 50 mM urea (Fig. 2). At present, we have no explanation for these divergent effects at low concentrations. However, there is no consistent evidence of special urea resistance in renal cells other than that which may result from accumulation of other organic solutes in vivo. Similar conclusions were reached in studies with marine cartilaginous fishes, which have 300- 500 mM urea throughout their intra- and extracellular fluids. These fishes have apparently evolved few urea- adapted proteins and instead appear to rely to a great extent on counteraction by methylamines (23, 24).
Betaine. This methylamine compound, found in high quantities in the inner mammalian kidney along with high urea concentrations, clearly penetrated MDCK cells and may even be accumulated above chemical equilib- rium (Table 1). In accompaniment with this, in the colony experiments, was a stimulation or lack of inhibi- tion of cell survival at low concentrations for all cell types (Figs. l-3), with a steep inhibition at concentra- tions ~50-100 mM for MDCK cells (Fig. 1A). These results are different from effects seen in experiments on isolated proteins, in which increases in thermostability
OPPOSING EFFECTS OF UREA AND BETAINE IN CELLS R203
or changes in Vmax or K, values are often linear with methylamine concentrations (23, 24, 27). However, al- though the stimulation of MDCK survival at 50 mM betaine is consistent with the counteracting osmolytes hypothesis, the inhibition at higher concentrations does not contradict it. There are at least two possible expla- nations. First, the inhibition may be a simple osmotic effect: although betaine clearly penetrated MDCK cells by the end of the 5-day growth period in the non-colony experiments (Table l), it may have crossed the mem- brane slowly and subjected the cell to high external osmolalities for several hours or days. This would not occur in vivo, where tissues are not exposed to rapid changes to high extracellular betaine. This osmotic ex- planation is supported by the steep inhibition curves seen for MDCK cells with NaCl and sorbitol (Fig. 1; Table 2), which are also not likely to cross cell mem- branes rapidly (see below). In addition, survival of both bacteria and halotolerant plants in hyperosmotic (NaCl) solutions is directly correlated with the ability to accu- mulate betaine (11,18). In such studies, betaine is termed a compatible osmolyte rather than a counteracting one because no urea is present. The lack of apparent com- patibility of betaine with MDCK cells at higher concen- trations could be a transmembrane osmotic effect caused by slow accumulation of betaine. Betaine would later act as a compatible solute once it was intracellular.
Second, betaine may rapidly cross the membrane and exert changes directly on intracellular macromolecular processes. Increases in extracellular osmotic strength have been shown to trigger active uptake in MDCK cells of betaine from the medium (E-17). In the present study, betaine itself may provide that osmotic signal. This is suggested by the fact that betaine accumulation is esti- mated to exceed chemical equilibrium (see RESULTS; Table 1). Inhibition of cellular processes by betaine could be consistent with the counteracting osmolytes hypoth- esis, even though methylamines alone (even at high concentrations) usually activate isolated proteins. As noted earlier, methylamines may inhibit some enzymes that urea activates (27)) increased structural stability may make some proteins too rigid for optimal function, and increased Vmax or decreased K, values may upset intricate cellular regulatory interactions absent in stud- ies with individual proteins (23, 24).
At present, we cannot readily distinguish between this and the first osmotic pressure hypothesis for the effects of betaine alone. A combination of effects may be occur- ring, for example, with stimulation of or compatibility with proteins dominating at low concentrations and os- motic effects dominating at higher concentrations.
NaCL. This salt was greatly inhibitory toward colony- forming efficiency and colony sizes with MDCK cells (Figs. IA and 4A). Although effects on macromolecular functions may be involved, it is likely that these are largely osmotic effects, since added NaCl does not cause any large sustained increase in intracellular NaCl in MDCK cells (15).
With PAP-HT25 and CHO cells, the lack of NaCl inhibition (Fig. 2) is explained by the ability of these cells to synthesize intracellular sorbitol in proportion to
extracellular osmolality, maintaining osmotic balance with a compatible osmolyte (3, 12). MDCK cells do not synthesize sorbitol in response to increased external osmotic pressure (15).
Polyols. Glycerol, the archetypical compatible solute used by a variety of organisms (6,23), surprisingly inhib- its colony-forming efficiency with MDCK cells but the extent was not as great as with urea or betaine at higher concentrations or other solutes at any concentration (Table 2 and Fig. 1A ). Also, glycerol had the least effect on colony size of all solutes tested (Fig. 4). Thus, overall, it is the most compatible solute of those tested, as might be expected from a variety of previous studies demon- strating its nonperturbing interactions with protein sys- tems (6, 23). The inhibition it did exhibit probably is not due to osmotic effects, since it is known to cross cell membranes readily (1). More likely, the inhibition may be caused by permeabilization of the cell membrane, a nonosmotic effect known to occur with some mammalian cells (1).
The other polyol tested, sorbitol, is also classified as a compatible solute based on its weak or stabilizing effects on proteins (19), although in one study sorbitol inhibited an enzyme far more than did glycerol (6). However, sorbitol inhibited colony-forming efficiencies of MDCK cells to the same extent as NaCl (Table 2), whose intra- cellular concentration rises little with extracellular con- centration (15). Although sorbitol clearly equilibrated across the cell membrane by the end of the experiments (several days) (Table l), the inhibition may simply be an osmotic effect. The sorbitol permeability of PAP-HT25 cells is low, -2 x lo-’ cm/s, and a full day is required for an amount of sorbitol equal to the cell content to leak out under hyperosmotic conditions (4). The equilibration between extracellular and intracellular fluid is presum- ably equally slow. Urea, on the other hand, equilibrates across the cell membranes of MDCK cells in ~30 min (1% .
Effect of Solute Combinations
The ability of betaine to offset the effects of urea on isolated proteins (23, 24) is mirrored by the present results with three distinct types of mammalian cells, including a nonrenal line (Figs. l-3). Addition of betaine to growth media with urea resulted in increased colony survival in almost every case. In addition to the penetra- tion of cells by both solutes (Table l), two patterns support the idea that this protection results from mac- romolecular interactions proposed by the counteracting osmolytes hypothesis. First, the ability of betaine to increase colony-forming efficiency in the presence of an inhibitory solute is largely confined to combinations with urea. Betaine had no effect on inhibition of colony- forming efficiency with cyanate (see MATERIALS AND METHODS) or glycerol (Fig. lC), and although it did raise colony-forming efficiency in the presence of 25 mM NaCl, it did not with 50 mM NaCl (Fig. 1C). This may be explained by membrane or osmotic effects of glycerol, osmotic effects of NaCl (>50 mosM), and by specific covalent modification of proteins by cyanate (see MATE- RIALS AND METHODS). Betaine would not be expected to
R204 OPPOSING EFFECTS OF UREA AND BETAINE IN CELLS
counteract these effects. Second, the polyols (glycerol, sorbitol) also penetrated the cells to near equilibrium but showed none of the counteracting properties in combi- nations with urea (Table 2) as seen with betaine. Again, the protective effect seems specific to the latter.
The concentrations used in this study approximate the range found in vivo for at least some mammalian kidneys. In rabbits, for example, urea in the inner medulla ranged from 50 to 350, betaine from 20 to 40, and another methylamine, glycerophosphorylcholine, from 10 to 40 mmol/kg wet wt in diuretic and antidiuretic states, re- spectively (25). Because urea is likely to be equilibrated intra- and extracellularly, while the methylamines are probably confined intracellularly (5), estimates of in vivo concentration ratios approach -2:l urea to methyla- mines (5, 26). Thus our results are consistent with the hypothesis that methylamines provide substantial tection from the inhibitory effects of urea in vivo.
pro-
We thank Ed Sokoloski of the National Heart, Lung, and Blood Institute for his analysis of growth media for cyanate.
Present address of P. H. Yancey: Biology Dept., Whitman College, Walla Walla, WA 99362.
Address for reprint requests: M. B. Burg, Bldg. 10, Rm. 6N307, National Institutes of Health, Bethesda, MD 20892.
Received 24 May 1989; accepted in final form 1 September 1989.
REFERENCES
1. ARMITAGE, W. J., AND P. MAZUR. Toxic and osmotic effects of glycerol on human granulocytes. Am. J. Physiol. 247 (Cell Physiol. 16): C382-C389,1984.
2. BAGNASCO, S., R. BALABAN, H. FALES, Y.-M. YANG, AND M. BURG. Predominant osmotically active organic solutes in rat and rabbit renal medullas. J. Biol. Chem. 261: 5872-5877, 1986.
3. BAGNASCO, S., S. UCHIDA, R. BALABAN, P. KADOR, AND M. BURG. Induction of aldose reductase and sorbitol in renal inner medullary cells by elevated extracellular NaCl. Proc. Nutl. Acad. Sci. USA 84: 1718-1720,1987.
4. BAGNASCO, S. M., H. R. MURPHY, J. J. BEDFORD, AND M. B. BURG. Osmoregulation by slow changes in aldose reductase and rapid changes in sorbitol flux. Am. J. Physiol. 254 (Cell Physiol. 23): C788-C792,1988.
5. BALABAN, R., AND M. BURG. Osmotically active organic solutes in the renal inner medulla. Kidney Int. 31: 562-564, 1987.
6. BROWN, A. D. Compatible solutes and extreme water stress in eukaryotic mica-organisms. Adu. Micro. Physiol. 17: 181-242,1978.
7. CHAMBERS, S. T., AND C. M. KUNIN. The osmoprotective proper- ties of urine for bacteria: the protective effect of betaine and human urine against low pH and high concentrations of electrolytes, sugars and urea. J. Infect. Dis. 152: 1308-1315, 1985.
8. CLARK, M. The osmotic role of amino acids: discovery and function. In: Transport Processes, Iono- and Osmoregulation, edited by R. Gilles and M. Gilles-Baillien. Berlin: Springer-Verlag, 1985, p. 412.
9. CSONKA, L. N. Physiological and genetic responses of bacteria to
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
osmotic stress. Microbial. Reu. 53: 121-147, 1989. GOTTESMAN, M. M. Growth properties of Chinese hamster ovary (CHO) cells. In: Molecular CeZl Genetics, edited by M. M. Gottes- man. New York: Wiley, 1985, p. 139-154. HIGGINS, F. H., J. CAIRNEY, D. A. STIRLING, L. SUTHERLAND, AND I. R. BOOTH. Osmotic regulation of gene expression: ionic strength as an intracellular signal? Trends Biol. Sci. 12: 339-344, 1987. HOHMAN, T., J. MILLEN, M. BOCK, AND D. DVORNICK. Hyperton- icity induces aldose reductase in cultured mesangial, endothelial and CHO cells (Abstract). J. Cell Biol. 107: 203a, 1988. LOW, P. Molecular basis of the biological compatibility of nature’s osmolytes. In: Transport Processes, Iono- and Osmoregulation, ed- ited by R. Gilles and M. Gilles-Baillien. Berlin: Springer-Verlag, 1985, p. 469. MARIER, J. R., AND D. ROSE. Determination of cyanate, and a study of its accumulation in aqueous solutions of urea. Anal. Biochem. 7: 304-314,1974. NAKANISHI, T., R. S. BALABAN, AND M. B. BURG. Survey of osmolytes in renal cell lines. Am. J. Physiol. 255 (Cell Physiol. 24): C181-C191,1988. NAKANISHI, T., AND M. B. BURG. Osmoregulatory fluxes of myo- inositol and betaine in renal cells. Am. J. Physiol. 257 (Cell Physiol. 26): C9644970,1990. NAKANISHI, T., R. J. TURNER, AND M. B. BURG. Osmoregulation of betaine transport in mammalian renal medullary cells. Am. J. Physiol. 258 (Renal Fluid Electrolyte Physiol. 27). In press. POLLARD, A., AND R. G. WYN JONES. Enzyme activities in concen- trated solutions of glycine betaine and other solutes. Planta 144: 291-298, 1979. SHIGRIN, S., AND C. L. PARROT. Influence of glycerol and other polyhydric alcohols on the quaternary structure of an oligomeric protein. Arch. Biochem. Biophys. 166: 426-432, 1975. STARK, G. R. Modifications of proteins with cyanate. In: Methods in Enzymology, edited by C. H. W. Hirs and S. N. Timasheff. New York: Academic, 1972, vol. 25, p. 579-584. UCHIDA, S., N. GREEN, H. COON, T. TRICHE, S. MIMS, AND M. BURG. High NaCl induces stable changes in phenotype and kary- otype of renal cells in culture. Am. J. Physiol. 253 (Cell Physiol. 22): C230-C242,1987. WOLFF, S. D., P. H. YANCEY, T. S. STANTON, AND R. S. BALABAN. A simple HPLC method for quantitating the major organic solutes of the renal medulla. Am. J. Physiol. 256 (Renal Fluid Electrolyte Physiol. 25): F954-F956, 1989. YANCEY, P., M. CLARK, S. HAND, R. BOWLUS, AND G. SOMERO. Living with water stress: evolution of osmolyte systems. Science Wash. DC 217: 1214-1222,1982. YANCEY, P. H. Organic osmotic effecters in marine cartilaginous fishes. In: Transport Processes, Iono- and Osmoregulation, edited by R. Gilles and M. Gilles-Baillien. Berlin: Springer-Verlag, 1985. YANCEY, P. H. Osmotic effecters in kidneys of xeric and mesic rodents: corticomedullary distributions and changes with water availability. J. Comp. Physiol. 158: 369-380, 1988. YANCEY, P. H., M. B. BURG, AND S. M. BAGNASCO. Effects of NaCl, glucose, and aldose reductase inhibitors on cloning efficiency of renal medullary cells. Am. J. Physiol. 258 (Cell Physiol. 27): C156-C163,1990. YANCEY, P. H., AND G. N. SOMERO. Methylamine osmoregulatory solutes of elasmobranch fishes counteract urea inhibition of en- zvmes. J. EXD. 2001. 212: 205-213. 1980.
